Fuzzy AI in General

Some work I have done:

JClips
SWT

Hmm... somehow i think I only mused in that post... ah well


AI or artificial intelligence is on the decline.

Or at least not as hot as it once was.

In the beginning AI languages were touted as the 5th generation language, after 4th generation languages.

The generation thing goes as follows:
1st gen: Machine Language baby! 01010110??? wtf????

2nd gen: Assembly Language - baby gimme an aspirin.. no wait mov that load

3rd gen: C mother of portability.... code on any machine baby (assembly still relies on ISA, ??? I cannot remember what that acronym stands for =P)

4th gen: ermmm huh I cannot remember but things like java??

5th gen: Oooh baby my code can code itself yippie!!! or something like that



As it turns out the closest we have at the moment are IDEs that make you stupid by automating EVERYTHING for you.

Well not really.

There is some AI activity going on, just that it lost its hype.


...



But I think it might be making a come back.


Now alot of people confuse knowledge management and knowledge engineering.


Knowledge Management is a methodology.
(actually it is just a fancy management term lol)

Essentially it has 4 parts which I cannot recall.
1) Knowledge Capturing
2) Knowledge Storage
3) Knowledge Retrieval
4) See I told you I cannot remember.. but I think it is transmission or something.



But take out all the fancy structure, and it just how do you capture and retrieve and disseminate information (or knowledge).

I think it is a worthy cause as mankind has always been about accumulating knowledge. From the time Man graduated from word of mouth to scribbling on paper. To printing paper, to printing online. Knowledge (or more specifically information) dissemination and accumulation has been around.

What does Knowledge Management offer then? KM tries to be more specify and anal about the knowledge part. How do you capture the "know how", the knowledge body that is stored in an employee's head (this is scary is'nt it... its like the matrix.. when they dehumanised a human bean into a battery).

So IBM has its red books and other players had other things. I think everything is still in flux nd eventually something will pan out. TThe problem with KM is twofold and it is a bit of a paradox.

KM suffers whether or not management decides to support it.


1) When Management does not support it:

"I need a Hero!" someone sang that in chicken little hoho. Yep, we need a champion.
But if that champion is alone with no management, it is gonna be hard. It is gonna be like Die Hard hard. An most of us ain't Bruce Willies. That champion would be a a tootless puppy with a loud bark and not much bite. Kinda like a terrier asking to be kicked. Now Open source ppl are different, so lets not consider those. We are talking about normal people in most organisations.


2) When management supports it:

Too often there is an agenda and management is doing it for some reason. The problem is that they are perfectly human, and by extension have a nose and by extension they have the proverbial nose to poke into the design, even though they have absolutely no clue what KM is. Its a double edged sword, everyone will want his way of implementing the KM. I think it really depends on the culture in the organisation. If it is command and control style, it is easy to get info captured, just bring the noose on the payslip. (not sure how that will affect staff morale though lol). On the other hand if your people are more accustomed to traditional pen and paper, that might be best. The key is: Whatever works.


Now that was my rant on KM. I don't think thats all but I am dry for now =P

Next up KE (Knowledge Enginneering) and how it is different from KM.




[to be cont']